The purpose of this solicitation is to create a database of literature citations which could be used to identify and classify grant applications by their scientific content. Specifically, the scientific bibliography (the "Literature Cited" section) of each application will be captured in machine-readable form. These data will then be put into a database with a prescribed format on the NIH mainframe computer. The contractor will generate the database essential for evaluating the bibliographic coupling method of indexing grant applications by their scientific content. If every incoming grant application could be classified (indexed) by its focus, NIH/DRG expects to streamline the process by which applications are assigned to study sections and funding components. In addition, the computer could be used to follow trends in application content to identify development of new fields or contraction of others, and to provide data to aid in making other decisions, such as creation, merging or dissolving study sections. With the current application volume, it is impractical for DRG to hire sufficient staff to index more than just the funded applications which are currently about 25% of the total submitted.